Oakland CPM Archive

home *** CD-ROM | disk | FTP | other *** search

/ Oakland CPM Archive / oakcpm.iso / cpm / zcpr33 / zarticls.lbr / ALIAS.AZT / ALIAS.ART

Wrap

Text File | 1988-02-25 | 13.5 KB | 378 lines

A STORY OF SUBMIT, EX, AND ZCPR3'S ALIAS, or How I Learned that ZCPR3 is the Pinnacle of CP/M by Rick Charnes June 12, 1986 [published in FOGHORN sometime around May-July 1987] Copyright (c) First Osborne Group and Rick Charnes. This articleáì may not be published anywhere without the express permission, ináì writing, from FOG. [This was my very first introduction to ZCPR3 - rc] ZCPR3 is great. I just found/invented/discovered the mostáì elegant way imaginable to deal with a particular application thatáì has challenged me for a year or so. A rather simple problem, butáì lots of fun to deal with. But let me go back to the beginning... Say I have a series of messages I've written to upload to aáì BBS. (Can you guess what I'm going to do with THIS message?áì Don't you just love the recursion?) First I have to "filter"áì them. I use Irv's FILTW. Now before EX15 came out which allowsáì Smartkey to load along with it (on top of it? below it?), Iáì couldn't use EX14 and so used plain old SUBMIT. My easy submitáì file was simply: áì ááááááááááFILTW $1 FILTW $2 FILTW $3 . . . FILTW $9 áì I rarely have 9 files that I need to filter...but this was justáì in case. On the command line I enter the filenames which getáì substituted as parameters, and away we go. If I only enter 3áì parameters, after the .SUB file has exhausted the 3, I hit ^C andáì the process is halted. Of course with SUBMIT you have to put upáì with silly and incessant disk accesses and the initial,áì laborious, and often noisy creation of the $$$.SUB file, etc. áì áì áááOK, then EX15 came along and most of the time Smartkey and Iáì were happy. Nice memory-based batch processing. But alas andáì alack--I couldn't get this particular .SUB file to work. In thisáì application--when I had a varying number of parameters on theáì command line and hence generally fewer than in the actual batcháì file, it just wouldn't budge. I would get the error message: No parameter or default parameter error line # 3 as in: A0>ex fil harddisk meáì EX 1.5 11-22-82 No parameter or default parameter error line # 3 filtw b:$3 and it simply wouldn't run. I found it somewhat ironic that aáì program as acknowledgedly more sophisticated than SUBMIT would beáì much pickier in insisting on an exact equation between the numberáì of parameters on the command line as in the command file. SUBMITáì didn't mind if there was a difference; you just hit CTL-C and itáì aborted. EX, however, wouldn't run at all. Now the documentation for EX that everyone has must be aáì classical case of how a programmer makes information as uniquelyáì inaccessible to the user as possible. I recognized that it hadáì something to do with what the author was calling "defaultáì parameter substitution." But for the life of me I couldn'táì figure out how to implement this feature; the exact syntaxáì completely eluded me. EX wanted to have parameters to stick ináì where I wasn't supplying them. As long as it had nothing, ratheráì than not having anything, it would be happy. <Heh-heh>. áì áááááI stared at the documentation until my eyes glazed over, andáì tried all sorts of different methods and couldn't get it to work. áì Finally I called Ken Fowler over at BAKUP, an excellentáì programmer, and with his help I was able to piece it together. áì the .SUB file actually has to have an initial line, differentáì from the actual command lines, looking precisely like: áì ááááá^$parm1 parm2 parm3 parm4 parm5 parm6 parm7 parm8 parm9 And I don't mean type the name of a parameter when you see theáì string "p-a-r-m" there; you actually type the letters p-a-r-m. áì EX takes it as some sort of null parameter name, I guess. All Iáì know is that it worked finally. "Worked finally" means here simply that I got it up to theáì level of SUBMIT, where I could enter any number of parametersáì from 1 to 9 on the command line and then abort the proceedingsáì with a CTL-C. Where FILTW.SUB is my submit/EX file, consistingáì of the above line as its first line, then the rest of the fileáì exactly as the SUBMIT file, and "ME" and "JU" are filenames, myáì terminal display now looked like this: A0>EX FILTW ME JU EX 1.5 11-22-82 (Ex Active) A0>filtw b:ME FILTW - Wordstar file filter Input and output files open Function complete. 7 input lines read 7 output lines with: 124 high-order bits deleted 0 CTL-characters deleted 0 dot commands removed 0 form feeds retained 0 soft hyphens fixed 0 space breaks fixed 0 orphan line feeds A0>filtw b:JU FILTW - Wordstar file filter Input and output files open Function complete. 26 input lines read 26 output lines with: 668 high-order bits deleted 0 CTL-characters deleted 0 dot commands removed 0 form feeds retained 0 soft hyphens fixed 0 space breaks fixed 0 orphan line feeds A0>filtw b:parm3 FILTW - Wordstar file filter ++ SOURCE FILE NOT FOUND ++ A0>filtw b:parm4 FILTW - Wordstar file filter ++ SOURCE FILE NOT FOUND ++ A0>filtw b:parm5 FILTW - Wordstar file filter ++ SOURCE FILE NOT FOUND ++ A0>filtw b:parm6 FILTW - Wordstar file filter ++ SOURCE FILE NOT FOUND ++ A0>filtw b:parm7 FILTW - Wordstar file filter >>>Ex Aborted<<< <--- this is where I hit CTL-C Now you may not think it such a victory to be able to get aáì program to give you 4 error messages and then abort, but for meáì it was. EX was substituting "parm" for the parameters, andáì finding no file, was passing on to the next, hoping to do betteráì this time. I suppose any string representing a filename thatáì doesn't exist would do; "p-a-r-m" really means nothing specificáì to EX. áì áì áááááI was partially satisfied and left the application and wentáì on to other things for a number of months. Until this eveningáì when I was riding home from work on BART and dreaming of ZCPR. Iáì was wondering if I could do something similar to the kind of workáì I had been doing with MexPlus. With MexPlus' advanced "script"áì file feature--amounting virtually to a programming language ofáì its own, leaving the public domain version in the dust looking atáì its own tail--I was writing files that would ask the user howáì many files he was uploading or transferring, then--using aáì looping/incrementing feature based on ye olde tried and true - áì ááááác=number of files to be transferred a=0 label begin a=a+1 . . . if a=c goto goodbye goto begin . . . label goodbye sendout "bye" ----MexPlus would then log off after the given number of filesáì had been transferred. So the question I was asking myself was:áì how could I get ZCPR to "test" a condition--how could I get it toáì determine how many parameters were entered and then stop byáì itself when it had processed all of them _but only that number?_ áì And furthermore, for the sake of clean and crisp programming, howáì could I create this with the absolute minimum ofáì characters/words/commands? áì "IF", "SETFILE", and "SYSTEM FILE #1" to the rescue. First to explain some of the individual components of Z3áì that I used. Number 1: the "SETFILE" command. Z3 uses what areáì called "system files". You can have up to 4 of them, and youáì define them with the SETFILE command. System files are justáì files that Z3 just keeps in its mind, ready to use them foráì whatever you want, by typing a particular string. See, Z3 has aáì better mind than CP/M. It can remember these things. If youáì tell Z3 "SETFILE 1 VDO.COM", then from then on--until I guess youáì cold boot--anytime you type "$N1" Z3 will substitute "VDO" rightáì in there and you can do with it whatever you please. You canáì imagine the extremely varied and powerful uses to which this canáì be put. It's just like a $n variable in SUBMIT except it'sáì hardier; it stays around from program to program, application toáì application. Anytime you want it, it's there just for theáì referencing. And again, you can define up to four of them. Individual component number 2: ZCPR3 has an "IF" commandáì which in turn has an "EXISTS" argument (can be shortened toáì "EX"), that does exactly the testing I want. If the conditionáì between it and the next "FI" (endif) is true---that is, if theáì following named file exists--then it continues on its merry way. áì If not, it aborts. Sounds just like what I need, eh? áì áááááAnd what ties it all together, number 3: ZCPR3's "ALIAS"áì command. An "ALIAS" is simply a .COM file that ZCPR makes,áì wrapping up any number of individual commands you give itáì (separated by semicolons) into a nice little package-in-one,áì processing them "en batch" when you give the single command. Anáì ALIAS is a perfect medium for the "system file" concept, and itáì will additionally allow normal "$1" variables a la SUBMIT _on topáì of that_. So naturally, I defined "FILTW" to be my system fileáì #1. My ALIAS, named FILTER.COM, now contains the followingáì commands, which do the job in the loveliest way: $N1 $1;IF EX $2;$N1 $2;FI;IF EX $3;$N1 $3;FI;IF EX $4;$N1 $4;FI Yes, I could have simply substituted "FILTW" where I have "$N1",áì but I wanted to have as much fun as possible. So: remember $N1áì expands to FILTW, and the parameters given on the command lineáì substitute for the normal $1, $2, etc. parameters. So, the firstáì thing this program (alias) does is to FILTW file1. Then itáì checks to see if parameter number 2 exists. If it doesn't,áì that's the end of that; abort. (Or, as Z3 enthusiasts put it,áì "set flow state to false.") Then, "FI" ends the "if" conditionáì testing. OK, start again: test to see if parameter 3 is there. áì If so, FILTW it; if not abort. Etc. Works like a charm; it'lláì do exactly what I want it do and no more. When it has exhaustedáì the parameters given it, it aborts back to the Z3 prompt andáì we're finished, ready now to send our messages to the BBS.áì One of the nicest parts of it for me is the watching and theáì seeing, or as some would say, the "progress reporting." It's aáì real delight to behold. You can see it process the "if"áì statements and go to the "fi" statements, and watch it settingáì the "command flow state" to true or false. Here's what I sawáì (all this courtesy of IO-CAP.COM, by the way, which _does_-- despite my initial skepticism--work under Z3): [My ALIAS isáì called FILTER.COM]: B:WORK>filter harddisk me me <-- I've entered 3 parameters FILTW - Wordstar file filter Input and output files open Function complete. 56 input lines read 56 output lines with: 0 high-order bits deleted 0 CTL-characters deleted 0 dot commands removed 0 form feeds retained 0 soft hyphens fixed 0 space breaks fixed 0 orphan line feeds IF T <----ZCPR3 has found the "if" condition (parameter 2áì ááááááááááááááexists) to be "T", that is, "true", so itáì áááááááááááááácontinues... FILTW - Wordstar file filter Input and output files open Function complete. 7 input lines read 7 output lines with: 0 high-order bits deleted 0 CTL-characters deleted 0 dot commands removed 0 form feeds retained 0 soft hyphens fixed 0 space breaks fixed 0 orphan line feeds To No IF <--- it's found the first "fi", but bypasses it IF T FILTW - Wordstar file filter Input and output files open Function complete. 7 input lines read 7 output lines with: 0 high-order bits deleted 0 CTL-characters deleted 0 dot commands removed 0 form feeds retained 0 soft hyphens fixed 0 space breaks fixed 0 orphan line feeds To No IFáì áIF F <---now the "if" condition is "f" "false", and after To No IF <--- this line, we're booted back to command level B:WORK>filter me <--- I tried it with only one parameter, toáì áááááááááááááááááááááááásee what fun could have FILTW - Wordstar file filter Input and output files open Function complete. 7 input lines read 7 output lines with: 0 high-order bits deleted 0 CTL-characters deleted 0 dot commands removed 0 form feeds retained 0 soft hyphens fixed 0 space breaks fixed 0 orphan line feeds IF F To No IF <---áì áIF F <--- I really like these guys; they're really neat. To No IF <--- Actually in real life they spread themselves áì áIF F <--- diagonally across the screen like trees ináì áTo No IF <--- bloom. [I discovered a year and thousands of adventures later that theseáì are the result of having the NOISE equate in SYSFCP.LIB set toáì TRUE. This apparently was the case with the early system I wasáì using. - rc, 2/23/88] B:WORK> All of the condition testing seems to happen in memory; thereáì are no disk accesses for the "if"s. It's very clean, quiet,áì fast, and extremely intelligent---just the way I like computersáì to be. áì And that's that. Z3 is a heck of a lot of fun. Why don'táì you give it a try? Rick Charnes "Home board" Morrow BBS in Oakland, California (415) 654-3882